The Role of Alloying Elements on the Fabricability of
Austenitic Stainless Steel
John C. Tverberg, P.E.
Metals and Materials Consulting Engineers
Mukwonago, Wisconsin
How many times have fabrication problems developed when a new coil or a new heat of
steel is put in production? The problems can be tearing, cracking, scratching, poorer
weld penetration, poor electropolished surface or a host of other problems. The usual
procedure to determine the source of the problem is a hardness test, tensile test, and
metallographic cross section and to review the mill test reports. Sometimes the source of
the problem is spotted, but most often nothing out of the ordinary is found. In these cases
the problem lies in the composition of the steel even when the alloy is within the
specified composition of the steel.
Alloy Design
Austenitic stainless steels are designed to give corrosion resistance in many
environments, resistance to hydrogen and 885º F (475º C) embrittlement, good strength,
good ductility and low hardness. In its simplest form stainless steel is iron with 12%
minimum chromium. This is what makes stainless steel rust resistant and allows the
passive film to develop.
Stainless steel exists in three metallurgical conditions depending on composition and heat
treatment: ferritic, martensitic and austenitic. These names refer to the crystallographic
structure: ferrite is body-centered cubic, austenite is face-centered cubic and martensite is
a distorted tetragonal which is the distorted face-centered cubic structure being changed
into a body-centered structure. The characteristics of these structures are tabulated in
Table I and are illustrated in Figure 1.
Table I: Metallurgical Types and Structures
Metallurgical Type
Crystal Type
No. Of Atoms
Phase Name
Ferritic
Body-Centered Cubic
9
Alpha or delta ferrite
Austenitic
Face-Centered Cubic
14
Gamma Iron
Martensitic
Distorted Tetragonal
10-13
Quenched or
Tempered Martensite
Pure iron is body-centered cubic, existing from absolute zero to its melting point. As
certain elements are added a “gamma loop” or austenite is created. These elements are
carbon, chromium, nickel, manganese, tungsten, molybdenum, silicon, vanadium and
2
silicon. Of these elements nickel, manganese, chromium and carbon have the ability to
extend the gamma loop the farthest. It is the combination of nickel and chromium that
allows the austenitic stainless steel to be face-centered cubic from absolute zero to the
melting point. It is this gamma loop that differentiates the ferritic alloys from the
martensitic alloys. Martensitic stainless steels have a narrow chromium range, 14 – 18%,
and must contain carbon, because only in this range can pure austenite be formed on
heating, thus obtaining martensite on quenching. Ferritic alloys are either below 14% or
above 18%. Varying the alloying elements can modify the range. The most common
method is to keep the carbon content low.
The most common austenitic stainless steels are based on the 18-8 composition, 18% Cr
+ 8% Ni. If the steel has more than 8% nickel it is austenitic, if it has less nickel then it
becomes duplex, that is austenite with ferrite islands. At 5% nickel the structure is about
50% austenite, 50% ferrite, and below 3% it becomes all ferrite. Therefore 8% nickel is
the basis for the lowest cost austenitic stainless steel.
When alloying elements are added to the steel, they may take one of the positions in the
basic crystal. These are called substitutional alloys and the alloy remains single phase.
Other elements are small enough they can fit between the atoms and are called interstitial
elements and the alloy remains single phase. Other elements will combine to form their
own unique crystals and form certain phases. Still others act as dirt in the alloy and are
called inclusions.
In this paper we will limit our discussion to the austenitic stainless steels, although many
of the comments will apply to the other types as well. Table II lists the controlled
elements in ASTM A240 compositions for the most common austenitic stainless steels.
Many of the troublesome elements are not controlled, so it is essential that they be called
out in either a customer specification or on the purchase order. Do not assume the alloy
you buy is made to the mid range of each element. Since 1988 the steel mills have
instituted a program called “alloy shaving” that uses the minimum alloying elements that
will just prevent hot shortness and cold rolling cracking.
Closer control of the alloying elements is possible because of the development of the
Argon-Oxygen-Decarburization (AOD) refining process that produces very low carbon
and sulfur content in the steel. The use of AOD allows the extensive use of scrap metal.
In fact, some heats are nearly all remelted scrap metal. The only disadvantage is a
continual build-up of non-specified tramp elements like copper, boron and calcium.
These elements may cause problems and their presence may explain why two different
heats that have the “same” chemistry, at least according to the CMTR, act so differently.
3
Table II: ASTM Composition of Common Austenitic Stainless Steels
UNS
S30100
S30200
S30400
S30403
S30409
S30451
S30453
S30500
S30908
S31008
S31600
S31603
S31651
S31653
S31700
S31703
S31725
S32100
Alloy
301
302
304
304L
304H
304N
304LN
305
309S
310S
316
316L
316N
316LN
317
317L
317LM
321
C
0.15
0.15
0.08
0.030
0.04-0.10
0.08
0.030
0.12
0.08
0.08
0.08
0.030
0.08
0.030
0.08
0.030
0.030
0.08
Mn*
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
Si*
1.00
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
1.50
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
Cr
16.00-18.00
17.00-19.00
18.00-20.00
18.00-20.00
18.00-20.00
18.00-20.00
18.00-20.00
17.00-19.00
22.0-24.00
24.00-26.00
16.00-18.00
16.00-18.00
16.00-18.00
16.00-18.00
18.00-20.00
18.00-20.00
18.00-20.00
17.00-19.00
Ni
6.00-8.00
8.00-10.00
8.00-10.50
8.00-12.00
8.00-10.50
8.00-10.50
8.00-12.00
10.50-13.00
12.00-15.00
19.00-22.00
10.00-14.00
10.00-14.00
10.00-14.00
10.00-14.00
11.00-15.00
11.00-15.00
13.50-17.50
9.00-12.00
Mo
…
…
…
…
…
…
…
…
…
…
2.00-3.00
2.00-3.00
2.00-3.00
2.00-3.00
3.00-4.00
3.00-4.00
4.0-5.0
…
N*
0.10
0.10
0.10
0.10
…
0.10-0.16
0.10-1.16
…
…
…
0.10
0.10
0.10-0.16
0.10-0.16
0.10
0.10
0.20
0.10
Cu
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
S34700
347
0.08
2.00
0.75
17.00-19.00
9.00-13.00
…
…
…
Other
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
…
Ti5(C+N)
min, 0.70
Max
Cb10Cmin
1.00Max
*Maximum
Note: Sulfur is limited to 0.030 max and Phosphorous to 0.045
The Role of Delta Ferrite
Delta ferrite usually is controlled to prevent microcracks in stainless steel welds during
welding. The best chart to predict the delta ferrite is the Schaeffler-DeLong Diagram,
Figures 2 and 3. The Welding Research Council (WRC) Diagram is the newest chart to
predict delta ferrite, but does not appear to be as good as the Schaeffler-DeLong Diagram
for the purpose of predicting fabrication difficulties since it does not take manganese into
account. Figure 2 has the ASTM composition limits for Type 304L and Figure 3 the
limits for Type 316L superimposed with actual ferrite numbers from 172 heats. The
trend in the past several years has been to higher delta ferrite contents.
As described earlier, delta ferrite has a body-centered cubic structure. It during
solidification and is retained in the structure at room temperature. Heat treatment, which
involves holding at 1900 - 2000º F (1040 - 1100ºC) for at least 10 minutes at temperature,
will dissolve delta ferrite up to 7-8%. Above that level, some delta ferrite remains in the
structure as a second phase. Additional heat treatment or longer time at temperature has
little effect on reducing the ferrite effect when the content is this high.
To calculate the delta ferrite in an alloy heat, use the composition listed in the certified
mill test report (CMTR) and plug the values into the chromium and nickel equivalent
equations for the Schaeffler-DeLong. Delta ferrite is controlled by the chromium,
molybdenum, niobium (columbium), silicon, nickel, carbon, nitrogen, manganese and
copper. The higher the delta ferrite, the more transformation martensite is formed and the
lower the ductility.
The presence of delta ferrite causes a number of problems with the austenitic stainless
steels:
4
1. In the lower nickel austenitic alloys such as Types 301, 302, 304, and 316,
cold working causes martensite to form. Higher delta ferrite means more
martensite transformation. As the martensite increases, the ductility decreases
and the potential for fracture increases.
2. If secondary operations like Electropolishing are performed, the delta ferrite
will preferentially dissolve leaving a white or dull surface. This is especially
true for welded parts. If the delta ferrite is too high, the part is unsuitable for
service in this environment since it cannot be heat treated out.
3. In corrosive environments, especially if acid chlorides are present, the delta
ferrite will preferentially dissolve. If the delta ferrite is present in welds, the
weld will totally dissolve. If the delta ferrite is too high, the part is unsuitable
for service in this environment since it cannot be heat treated out.
4. If the delta ferrite is below 2% the alloy is good for fabrication, but the alloy
may develop microcracks during welding.
The optimum delta ferrite content for welding is 2 – 5% as determined by the Welding
Research Council. This should be the aim-for range for all of the low nickel austenitic
stainless steels.
The Role of Carbon
Since the advent of AOD refining, carbon control in stainless steel is almost a non-issue
except for those who specify the “straight” grades either through ignorance or for the
higher ASME allowable stress values from the Boiler and Pressure Vessel Code. The
“straight” grades are the original, pre-AOD refining era, high carbon varieties, usually
with 0.08, 0.12 or 0.15% C. For most applications the “L” grades, with 0.030% C
maximum, are preferred and should be used wherever possible.
Carbon is an interstitial alloying element. As a result it can diffuse rapidly through the
structure and concentrate on the grain boundaries. The primary problem is precipitation
of chromium carbide. Formation of chromium carbide depletes the grain boundaries of
chromium, thus destroying the corrosion resistance. Fortunately, a short-time heat
treatment at 1850º F (1010º C) will dissolve most chromium carbide, except when the
carbon content is so high the part cannot be quenched fast enough. Figure 4 illustrates
the effect of time-temperature and carbon content on the formation of chromium carbide
on cooling from heat treatment temperatures. This chart shows that the typical carbon
levels in today’s AOD refined “L” grade steels of 0.024% would require nearly 50 hours
at 1000º F (540º C) to cause chromium carbide to precipitate. This chart also indicates
that carbides will eventually form if the component is held at temperatures in the range of
1000º F (540º C). The solution is to use a stabilized stainless steel that contains either
titanium or niobium (columbium). These elements preferentially form carbides, thus
prevent chromium carbide from forming.
Carbon has a number of effects on the fabricability:
1. High carbon increases the tensile strength of the alloy and its hardness. With
this increased strength comes decreased ductility, thus a greater tendency for
fracture. It is not possible to anneal a high carbon stainless steel to a low
hardness level. The only solution is to use “L” grade alloys.
5
2. High carbon means difficulty during quenching after heat treatment. To
prevent carbide precipitation, the alloy must be quenched at or near the
solution anneal temperature. Some high carbon grades are impossible to
quench fast enough to prevent carbide precipitation, especially those with
heavy cross sections.
3. Field welding the straight grades, where it is not possible to heat treat after
welding, will corrode in even benign water.
Calcium
Calcium is a strong deoxidizer and is used by some steel mills to kill the melt. Most of
the calcium comes from the slag, refractories and mold powders during melting and
casting. It is associated with boron in some grain boundary precipitates as Figure 8
shows. These compounds form very slaggy welds. Most steels run in the range of
0.0010 – 0.0020%, but occasionally a very slaggy weld will have 0.0145 – 0.00175%
calcium. The lower limits do not appear to be a problem, but the higher levels are.
Boron
Boron is one of those elements that some steel mills love and other mills avoid. Boron is
added to the melt for sliver control during hot rolling. Boron is not an element that builds
up as a residual. It is easy to tell which mills use boron because their heats are in the
range of 0.0070 – 0.0185%. The solubility limit is reported to be in the range of 0.010%.
Steel from the mills that avoid the use of boron is in the range of >0.0001 – 0.0006%.
Boron is impossible to identify using microprobe analysis because of its low atomic
number. The only reliable analytical method is inductively coupled plasma spectroscopy
with the appropriate standards. Standard emission spectroscopy does not appear to detect
boron with the same sensitivity as ICP spectroscopy.
Sulfur
Weld penetration difficulty can occur when the sulfur content is too low. During AOD
refining the residual sulfur is reduced to 0.001% or less. This is desirable to produce low
inclusion steel, but very bad for weld penetration. In recent years the BPE (BioPharmaceutical Equipment) Committee of the ASME has specified the sulfur content of
all steel components for the pharmaceutical industry to have a sulfur content of 0.005 –
0.017%. This means the steel must be resulfurized at the time of casting. Most of these
steels have a mean sulfur range of 0.012 – 0.014%. No adverse problems have been
identified with sulfur at these low concentrations except a tendency to have a higher
inclusion count. Inclusions are insoluble non-metallic “dirt” in the steel. One school of
thought says these inclusions may lead to a higher pitting corrosion rate.
Sulfur, together with controlled manganese and silicon, control the convection currents in
the weld pool so the heat is carried deep into the weld. Elements that reverse the
convention currents producing low penetration welds include calcium, titanium,
aluminum, magnesium and the rare earths.
6
Manganese
Another element that aids weld penetration is manganese. The ideal range is between 1.5
and 1.75%. Manganese does three things to aid the stability of stainless steel:
• promotes the formation of austenite
• combines with the sulfur to form manganese sulfides
• increases the solubility of nitrogen.
The only negative aspect of manganese is its tendency to form inclusions in the steel,
especially in the presence of sulfur. For this reason some semiconductor applications
limit manganese to 0.5% maximum or in some cases to 0.05% maximum.
Silicon
Silicon is added to the steel to assist deoxidation following AOD refining and to improve
fluidity of the molten metal. Several problems can arise from high silicon levels during
welding:
• “sagging” of the weld;
• promote ferrite formation;
• and formation of weld slag spots, especially if the cover gas is contaminated
with oxygen or if moisture is present.
Most of the common austenitic stainless steels limit silicon at 0.75%. If the silicon
becomes too high it will promote ferrite formation, and during cold working form
transformation martensite. The presently accepted silicon range is 0.4 – 0.5%.
Copper
The jury is still out on copper. In the range of 3 – 4% copper improves corrosion
resistance to sulfuric acid. At low levels and in combination with other elements copper
will produce age hardening. But the copper content is creeping upward because it is
being introduced through the use of scrap metal during compounding the melt, and AOD
refining does not eliminate it. Today, the copper content of most of the common stainless
grades, 304, 304L, 316, 316L, is in the range of 0.4 – 0.5%. Ten years ago they were half
that. Table II shows that copper is not controlled. Copper may not be all-bad, since it
improves resistance to sulfuric acid corrosion. However, if the alloy is to be used for any
type of human implant it may cause problems.
Titanium and Zirconium
Titanium and/or zirconium are popular elements for deoxidizing the melt after AOD
refining. Deoxidation is necessary to prevent fracture of the ingot during hot rolling, so
the steel mills want totally “killed” steels. Because oxygen promotes weld penetration,
excess titanium and/or zirconium will reverse convection currents and cause variations in
weld uniformity. Longitudinal welds will develop arrowheads, a slag spot will form, and
then the weld becomes normal again. In addition, excess titanium and/or zirconium
scavenge oxygen from the cover gas and forms slag spots on the weld. It appears that
titanium/zirconium levels in the range of 0.001 – 0.004% have little effect on the weld
quality, but when the concentration is in excess of 0.007% problems may develop.
7
Aluminum
Aluminum is another element that is used to deoxidize the steel. It acts much the same as
titanium both as a slag former and its tendency to form arrowheads on the weld with an
associated slag spot. Aluminum in the range of 0.006% appears to be safe, but residuals
in excess of 0.010 may cause slag formation.
Magnesium
Like titanium and aluminum, magnesium is a strong deoxidizer. In the past it was used in
some deoxidizer compounds, but today it is normally not added to the melt for
deoxidization. It shows up in the steel from the slag and refractory lining of the AOD
converter, ladle and caster. All steel mills have about the same amount of residual
magnesium, in the order of 0.006 – 0.008%. This does not appear to be a problem.
Rare Earths
The common rare earths include cesium, yttrium and lanthanum. They are powerful
deoxidizers and result in lack of weld adherence of the joint surfaces and weld cracks.
Those alloys that have intentional additions are variable in weldability, exhibit weld cross
cracks and may show balling of the weld metal on the surface. The levels where
problems develop are in the range of 0.005% of any of the three elements. Thankfully,
very few mills use rare earth additions. None of them should.
Summary
Variations within the composition limits of stainless steel alloys can affect both the
mechanical working and welding characteristics of the alloy. These variations can cause
different heats of steel to perform quite differently. Today’s steel is melted to the lean
side of the composition resulting in problems that were not encountered 15 or more years
ago. In addition the tramp elements are increasing and changes in deoxidation practice
are introducing other problems. Specifically, these are the factors that affect fabricability
and weldability:
1.
Delta ferrite content is increasing and with it reduced ductility because
of martensite formation. Delta ferrite in welds cause corrosion
problems and pitted welds following Electropolishing. The elements
that control delta ferrite formation are chromium, molybdenum,
niobium (columbium), silicon, nickel, carbon, nitrogen, manganese and
copper.
2.
Carbon increases the strength and hardness of the steel ands decreases
the ductility. In addition, carbon above 0.035% may precipitate along
the grain boundaries during welding.
3.
Sulfur is desirable for welding since it promotes deep penetration.
However, it forms sulfide inclusions in the steel that may affect surface
quality, perhaps pitting and act as stress risers.
4.
Manganese promotes weld penetration but forms manganese sulfide
inclusions. Manganese increases the solubility of nitrogen, so it is of
help with high nitrogen steels.
5.
Silicon, in moderation, is good. In excess, silicon increases the delta
ferrite, causes welds to sag and become slaggy.
8
6.
7.
8.
9.
10.
11.
Copper is increasing in the steel because of remelting so much scrap.
The jury is still out on its effect.
Titanium and zirconium are added as deoxidizers or carbide formers. It
appears that a little is good, but too much is bad.
Aluminum is added as a deoxidizer, so a little is good. Too much
aluminum causes ductility and welding problems. The same is true for
magnesium.
Calcium is used as a deoxidizer and is picked up from the melting
refractories. It is bad.
Boron is used as a means of sliver control during hot rolling. It is bad.
Some mills add rare earths a deoxidizer. They cause problems during
welding. Except for use in certain high temperature superalloys they
have few redeeming qualities.
Body-Centered Cubic
Face-Centered Cubic
Figure 1: Atomic arrangement for the unit cells of stainless steel. Ferrite is BodyCentered Cubic and Austenite is Face-Centered Cubic.
9
Figure 2: The Shaeffler-DeLong Diagram for determining the delta ferrite potential based
on the alloy composition. This hart shows the composition limits for Type 304L. The
black dots represent actual heats. The group clustered around 5 – 6% is older heats, the
group around 10-12 FN represent the newer heats.
Figure 3: The Schaeffler-DeLong Diagram with the composition limits for 316L
superimposed. Note that most of the 72 compositions are clustered around 4% FN and
later heats around 6 FN. This delta ferrite range should not cause problems during
fabrication.
.
10
Figure 4: Effect of carbon on the time required to form chromium carbide. Carbide
precipitation occurs inside the loop to the right of various carbon content curves.
Figure 5 The effect of calcium boride on the ductility of Type 316L stainless steel. The
black areas are cracks, and the small gray areas around the grain boundaries, especially in
the lower right corner, are the calcium boride particles. They cannot be annealed out.
500X